Page 6 of 9

Re: Performance Benchmarks

Posted: Sun Jul 05, 2015 11:57
by t13swift
t13swift wrote:Custom Desktop
Intel i5 4460 3.2GHz
Nvidia GTX 760
16GB RAM
128GB SSD for OS
256GB SSD for Content
Windows 7
Arena 4.20

GUI @1080p, 2x1080p outputs

1080p clean 20 at 46fps
1080p noise 19 at 34fps

4K clean 9 at 31fps
4K noise 8 at 30fps
Tested the same rig with a 980Ti, Arena 4.2.1:
4K clean 12@36fps, 13@19fps
4K noise 11@32fps, 12@28fps

Re: Performance Benchmarks

Posted: Wed Aug 12, 2015 05:15
by dirtyjohn_lv
Got around to installing the Titan X in my main machine. Running this with a couple display port splitters and 1 GTX 780 (until my Datapath x4 and extra screen get installed). It does quite well.

The Titan X only supports 4 monitors, hence the splitters and extra card. For some reason, adding the second card will only allow 5 monitors (when connected to second card). Once I installed the SLI bridge (even though SLI is turned off) I can connect a 6th screen.

Model: Custom Desktop
Operating System: Win 7 64
GPU: NVIDIA Titan X 12GB
Harddisk: Samsung SSD 840, 512 GB x 4 (RAID 0)
Processor: Core i7 4930k 3.4 GHz
Memory: 64GB RAM
Resolume: 4.2.1

Output Setup:
1 Control Monitor at 1600x900 (Touchscreen),
1 LED @ 1440x900
3 LED @ 1280x720
1 BMD DVI Extender @ 1080i

Benchmark 1920x1080 DXV2
Minimal content: 20 layers @ 50fps
Noise content: 20 layers @ 31fps

Benchmark 1920x1080 DXV3
Minimal content: 20 layers @ 100fps
Noise content: 20 layers @ 45fps

Benchmark 4k DXV3
Minimal content: 12 layers @ 31fps
Noise content: 10 layers @ 31fps


My NVIDIA 3D settings in case anyone is curious
Ambient Occlusion: Performance
Anisotropic filtering: Application Controlled
Antialiasing - FXAA: On
Antialiasing - Gamma Correction: On
Antialiasing - Mode: Application Controlled
Antialiasing - Setting: Application Controlled (greyed out disabled)
Antialiasing - Transparency: Multisample
CUDA - GPUs: All
DSR - Factors: Off
DSR - Smoothness: Off (greyed out disabled)
Maximum pre-rendered frames: Use 3D application setting
Multi-Frame Sampled AA (MFAA): On
Multi-display/mixed-GPU acceleration: Multiple display performance mode
Power management mode: Prefer maximum performance
Shader Cache: On
Texture filtering - Anisotropic sample optimization: On
Texture filtering - Negative LOD bias: Allow
Texture filtering - Quality: High performance
Texture filtering - Trilinear optimization: On
Threaded optimization: Auto
Triple buffering: Off
Vertical sync: Off
Virtual Reality pre-rendered frames: 1 (This used to be just pre-rendered frames on previous cards)


Also under "Configure Surround, PhysX" I have the PhysX Processor set to the primary card (no Auto)

Also under "Adjust desktop size and position" I have every screen set to No scaling and "Perform scaling" set to GPU

Re: Performance Benchmarks

Posted: Thu Aug 13, 2015 08:11
by cosmowe
You turned Vertical sync: Off....so, do you have visible tearing?

greetings
cosmowe

Re: Performance Benchmarks

Posted: Thu Aug 13, 2015 09:25
by dirtyjohn_lv
Edit: Made a topic in General Discussion so we dont clutter up benchmark

viewtopic.php?f=5&t=12269



No visible tearing on any of the stuff I normally run, didn't really look to hard when running benchmark. I will run it again tomorrow and record my feed to my Tricaster for better analysis.

Turning Vsync off is part of the reason I get such high framerates. I have tried turning it on, but anything I load starts at 30fps

Re: Performance Benchmarks

Posted: Fri Sep 18, 2015 08:40
by Zoltán
dell inspiron 15, 5000 series
i7 5500u @ 2,4GHz
8GB RAM
output to 1080p.
1TB HDD

Intel HD graphics 5500
  • 1080p
    • noise 3 layers
      clean 3 layers
    480p
    • clean 19 layers
      noise 18 layers
Nvidia GeForce 920M
  • 1080p
    • noise 3 layers
      clean 8 layers
    480p
    • clean 20+ layers, 20layers @54 fps
      noise 20 layers @ 32 fps

Re: Performance Benchmarks

Posted: Fri Oct 16, 2015 13:16
by cosmowe
Brand new R5...new Laptop.......so it's time for a new Benchmark run.

Laptop: XMG Schenker U506
CPU: i7 6700K Skylake desktop CPU 4,2 Ghz
GPU: GTX 970M 6GB
RAM: 16 GB DDR4 2133 MHz
SSD: 512GB m.2 SSD Samsung SM951 PCI-Express x4
OS: Windows 10
SOFT: Resolume Arena 5.0.0

Output Setup: 1x GUI at 1920 x 1080 60Hz, 1x Fullscreen Output at 1920x1080 60Hz

1080p Benchmark
Minimal content DXV3: 20+ layers (20 @ 60 fps)
Noise content DXV3: 20+ layers (20 @ 55 fps)

4K Benchmark
Minimal content DXV3: 11 layers (11 @ 31 fps)
Noise content DXV3: 8 layers (8 @ 30 fps)

Extra info: Same results with R4 4.2.2

This small and light notebook went quite well I guess...*happy*...I can really recommend this XMG notebook especially because of the 3 video outputs. 2x Fullsize Displayport 1x HDMI


Greetings
cosmowe

Re: Performance Benchmarks

Posted: Fri Oct 16, 2015 13:38
by Oaktown
Hello Cosmowe.

Very impressive!

Do you have any way to test with 1 or 2 4K output(s)?

Re: Performance Benchmarks

Posted: Fri Oct 16, 2015 13:53
by cosmowe
Hi Oaky ;)
nope...not yet.

submissively
cosmowe

Re: Performance Benchmarks

Posted: Fri Oct 16, 2015 15:26
by Oaktown
What! No Datapath X4 units in the top drawer of your A/V cabinet! ;)

What did you pay for your little wonder? I read the review and it really sounds promising.

What made you choose the 970M 6GB over the 980M 8GB?

Re: Performance Benchmarks

Posted: Fri Oct 16, 2015 23:05
by cosmowe
What! No Datapath X4 units in the top drawer of your A/V cabinet! ;)
No.....till now I don't need it.
What did you pay for your little wonder? I read the review and it really sounds promising.
2400€
What made you choose the 970M 6GB over the 980M 8GB?
Simply the price....my goal was to refund the notebook with a single job. And in mind that the technology will show up with something much more powerfull in two years than the 980m and the i7 skylake. How many 4k jobs do you have in a year?......I had one in 2015 which really really needs a 4 k output.....and i solved it with a pandoras box....... So.....thats the reason.
There will be much more in the future.....but at last in the vj business. I try to think a little bit about the money.